Defective chemical process chain detection in uncertainty condition

An approach to the building of a hierarchical hybrid expert system for the technical diagnosis of chemical plants is presented that utilizes both qualitative and quantitative process knowledge. Evidential intervals are used to make allowances for the inaccuracy of expert estimations, instability of...

Full description

Saved in:
Bibliographic Details
Published in:Computers & chemical engineering 1993, Vol.17 (10), p.1015-1024
Main Authors: Perov, V.L., Paliukh, B.V., Kuperman, V.G.
Format: Article
Language:English
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:An approach to the building of a hierarchical hybrid expert system for the technical diagnosis of chemical plants is presented that utilizes both qualitative and quantitative process knowledge. Evidential intervals are used to make allowances for the inaccuracy of expert estimations, instability of parameters, measurement noise etc. Diagnostic variables are represented as interval numbers. It is shown that such an approach gives a number of advantages over Bayesian and fuzzy set techniques. A procedure for identification of possibly faulty process chains in the first stage of a diagnosis is described using the example of a nitric acid production process. Detailed diagnosis, not treated in this paper, would be applied after candidate process chains are located.
ISSN:0098-1354
1873-4375
DOI:10.1016/0098-1354(93)80082-X